gsm8k-lora / trainer_state.json
HongzheBi's picture
Upload 8 files
5eacd2c
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.7198549410698094,
"global_step": 3000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.01,
"learning_rate": 4.999908736188902e-05,
"loss": 1.3169,
"step": 10
},
{
"epoch": 0.02,
"learning_rate": 4.999593264902687e-05,
"loss": 1.0577,
"step": 20
},
{
"epoch": 0.03,
"learning_rate": 4.999052487856493e-05,
"loss": 0.9159,
"step": 30
},
{
"epoch": 0.04,
"learning_rate": 4.998286453794323e-05,
"loss": 0.8264,
"step": 40
},
{
"epoch": 0.05,
"learning_rate": 4.9972952317641664e-05,
"loss": 0.7412,
"step": 50
},
{
"epoch": 0.05,
"learning_rate": 4.996078911111772e-05,
"loss": 0.6837,
"step": 60
},
{
"epoch": 0.06,
"learning_rate": 4.994637601472596e-05,
"loss": 0.7422,
"step": 70
},
{
"epoch": 0.07,
"learning_rate": 4.99297143276192e-05,
"loss": 0.6616,
"step": 80
},
{
"epoch": 0.08,
"learning_rate": 4.9910805551631424e-05,
"loss": 0.7062,
"step": 90
},
{
"epoch": 0.09,
"learning_rate": 4.988965139114237e-05,
"loss": 0.657,
"step": 100
},
{
"epoch": 0.1,
"learning_rate": 4.986625375292394e-05,
"loss": 0.6791,
"step": 110
},
{
"epoch": 0.11,
"learning_rate": 4.984061474596833e-05,
"loss": 0.6274,
"step": 120
},
{
"epoch": 0.12,
"learning_rate": 4.9812736681297904e-05,
"loss": 0.6117,
"step": 130
},
{
"epoch": 0.13,
"learning_rate": 4.978262207175689e-05,
"loss": 0.63,
"step": 140
},
{
"epoch": 0.14,
"learning_rate": 4.9750273631784924e-05,
"loss": 0.6091,
"step": 150
},
{
"epoch": 0.15,
"learning_rate": 4.97227884745494e-05,
"loss": 0.633,
"step": 160
},
{
"epoch": 0.15,
"learning_rate": 4.968642662273368e-05,
"loss": 0.6445,
"step": 170
},
{
"epoch": 0.16,
"learning_rate": 4.9647839611253184e-05,
"loss": 0.6469,
"step": 180
},
{
"epoch": 0.17,
"learning_rate": 4.960703091822416e-05,
"loss": 0.588,
"step": 190
},
{
"epoch": 0.18,
"learning_rate": 4.95640042220185e-05,
"loss": 0.5695,
"step": 200
},
{
"epoch": 0.19,
"learning_rate": 4.9518763400932166e-05,
"loss": 0.6091,
"step": 210
},
{
"epoch": 0.2,
"learning_rate": 4.9471312532835645e-05,
"loss": 0.6153,
"step": 220
},
{
"epoch": 0.21,
"learning_rate": 4.942165589480634e-05,
"loss": 0.6056,
"step": 230
},
{
"epoch": 0.22,
"learning_rate": 4.936979796274309e-05,
"loss": 0.5868,
"step": 240
},
{
"epoch": 0.23,
"learning_rate": 4.931574341096269e-05,
"loss": 0.604,
"step": 250
},
{
"epoch": 0.24,
"learning_rate": 4.925949711177859e-05,
"loss": 0.5964,
"step": 260
},
{
"epoch": 0.24,
"learning_rate": 4.9201064135061694e-05,
"loss": 0.5506,
"step": 270
},
{
"epoch": 0.25,
"learning_rate": 4.914044974778341e-05,
"loss": 0.602,
"step": 280
},
{
"epoch": 0.26,
"learning_rate": 4.907765941354085e-05,
"loss": 0.5943,
"step": 290
},
{
"epoch": 0.27,
"learning_rate": 4.9012698792064426e-05,
"loss": 0.565,
"step": 300
},
{
"epoch": 0.28,
"learning_rate": 4.894557373870764e-05,
"loss": 0.5818,
"step": 310
},
{
"epoch": 0.29,
"learning_rate": 4.887629030391932e-05,
"loss": 0.6095,
"step": 320
},
{
"epoch": 0.3,
"learning_rate": 4.880485473269828e-05,
"loss": 0.5539,
"step": 330
},
{
"epoch": 0.31,
"learning_rate": 4.873127346403035e-05,
"loss": 0.5863,
"step": 340
},
{
"epoch": 0.32,
"learning_rate": 4.865555313030805e-05,
"loss": 0.5869,
"step": 350
},
{
"epoch": 0.33,
"learning_rate": 4.857770055673274e-05,
"loss": 0.5819,
"step": 360
},
{
"epoch": 0.34,
"learning_rate": 4.849772276069941e-05,
"loss": 0.6066,
"step": 370
},
{
"epoch": 0.34,
"learning_rate": 4.841562695116415e-05,
"loss": 0.5623,
"step": 380
},
{
"epoch": 0.35,
"learning_rate": 4.833142052799439e-05,
"loss": 0.5557,
"step": 390
},
{
"epoch": 0.36,
"learning_rate": 4.824511108130186e-05,
"loss": 0.5869,
"step": 400
},
{
"epoch": 0.37,
"learning_rate": 4.815670639075845e-05,
"loss": 0.5145,
"step": 410
},
{
"epoch": 0.38,
"learning_rate": 4.8066214424895e-05,
"loss": 0.5519,
"step": 420
},
{
"epoch": 0.39,
"learning_rate": 4.7973643340383004e-05,
"loss": 0.6257,
"step": 430
},
{
"epoch": 0.4,
"learning_rate": 4.7879001481299404e-05,
"loss": 0.5976,
"step": 440
},
{
"epoch": 0.41,
"learning_rate": 4.778229737837451e-05,
"loss": 0.585,
"step": 450
},
{
"epoch": 0.42,
"learning_rate": 4.768353974822301e-05,
"loss": 0.6175,
"step": 460
},
{
"epoch": 0.43,
"learning_rate": 4.758273749255833e-05,
"loss": 0.6172,
"step": 470
},
{
"epoch": 0.44,
"learning_rate": 4.7479899697390235e-05,
"loss": 0.5003,
"step": 480
},
{
"epoch": 0.44,
"learning_rate": 4.737503563220586e-05,
"loss": 0.5583,
"step": 490
},
{
"epoch": 0.45,
"learning_rate": 4.726815474913417e-05,
"loss": 0.5686,
"step": 500
},
{
"epoch": 0.46,
"learning_rate": 4.715926668209398e-05,
"loss": 0.5621,
"step": 510
},
{
"epoch": 0.47,
"learning_rate": 4.704838124592558e-05,
"loss": 0.558,
"step": 520
},
{
"epoch": 0.48,
"learning_rate": 4.693550843550607e-05,
"loss": 0.5746,
"step": 530
},
{
"epoch": 0.49,
"learning_rate": 4.682065842484843e-05,
"loss": 0.5191,
"step": 540
},
{
"epoch": 0.5,
"learning_rate": 4.670384156618448e-05,
"loss": 0.5243,
"step": 550
},
{
"epoch": 0.51,
"learning_rate": 4.658506838903176e-05,
"loss": 0.5996,
"step": 560
},
{
"epoch": 0.52,
"learning_rate": 4.6464349599244425e-05,
"loss": 0.5799,
"step": 570
},
{
"epoch": 0.53,
"learning_rate": 4.634169607804826e-05,
"loss": 0.5152,
"step": 580
},
{
"epoch": 0.53,
"learning_rate": 4.6217118881059864e-05,
"loss": 0.5641,
"step": 590
},
{
"epoch": 0.54,
"learning_rate": 4.609062923729014e-05,
"loss": 0.5278,
"step": 600
},
{
"epoch": 0.55,
"learning_rate": 4.596223854813219e-05,
"loss": 0.5245,
"step": 610
},
{
"epoch": 0.56,
"learning_rate": 4.5831958386333505e-05,
"loss": 0.503,
"step": 620
},
{
"epoch": 0.57,
"learning_rate": 4.569980049495298e-05,
"loss": 0.5509,
"step": 630
},
{
"epoch": 0.58,
"learning_rate": 4.5565776786302336e-05,
"loss": 0.5467,
"step": 640
},
{
"epoch": 0.59,
"learning_rate": 4.5429899340872405e-05,
"loss": 0.5637,
"step": 650
},
{
"epoch": 0.6,
"learning_rate": 4.529218040624425e-05,
"loss": 0.5836,
"step": 660
},
{
"epoch": 0.61,
"learning_rate": 4.5152632395985165e-05,
"loss": 0.5497,
"step": 670
},
{
"epoch": 0.62,
"learning_rate": 4.5011267888529786e-05,
"loss": 0.5182,
"step": 680
},
{
"epoch": 0.63,
"learning_rate": 4.486809962604631e-05,
"loss": 0.5251,
"step": 690
},
{
"epoch": 0.63,
"learning_rate": 4.472314051328794e-05,
"loss": 0.5453,
"step": 700
},
{
"epoch": 0.64,
"learning_rate": 4.457640361642968e-05,
"loss": 0.5442,
"step": 710
},
{
"epoch": 0.65,
"learning_rate": 4.442790216189062e-05,
"loss": 0.5223,
"step": 720
},
{
"epoch": 0.66,
"learning_rate": 4.4277649535141726e-05,
"loss": 0.5733,
"step": 730
},
{
"epoch": 0.67,
"learning_rate": 4.412565927949933e-05,
"loss": 0.5234,
"step": 740
},
{
"epoch": 0.68,
"learning_rate": 4.397194509490433e-05,
"loss": 0.5091,
"step": 750
},
{
"epoch": 0.69,
"learning_rate": 4.381652083668739e-05,
"loss": 0.506,
"step": 760
},
{
"epoch": 0.7,
"learning_rate": 4.3659400514320016e-05,
"loss": 0.5328,
"step": 770
},
{
"epoch": 0.71,
"learning_rate": 4.3500598290151796e-05,
"loss": 0.5134,
"step": 780
},
{
"epoch": 0.72,
"learning_rate": 4.3340128478133834e-05,
"loss": 0.5423,
"step": 790
},
{
"epoch": 0.73,
"learning_rate": 4.317800554252857e-05,
"loss": 0.5101,
"step": 800
},
{
"epoch": 0.73,
"learning_rate": 4.301424409660598e-05,
"loss": 0.5427,
"step": 810
},
{
"epoch": 0.74,
"learning_rate": 4.284885890132639e-05,
"loss": 0.526,
"step": 820
},
{
"epoch": 0.75,
"learning_rate": 4.2681864864009985e-05,
"loss": 0.5381,
"step": 830
},
{
"epoch": 0.76,
"learning_rate": 4.251327703699308e-05,
"loss": 0.5581,
"step": 840
},
{
"epoch": 0.77,
"learning_rate": 4.2343110616271376e-05,
"loss": 0.5464,
"step": 850
},
{
"epoch": 0.78,
"learning_rate": 4.217138094013021e-05,
"loss": 0.5458,
"step": 860
},
{
"epoch": 0.79,
"learning_rate": 4.1998103487762034e-05,
"loss": 0.5158,
"step": 870
},
{
"epoch": 0.8,
"learning_rate": 4.1823293877871164e-05,
"loss": 0.5544,
"step": 880
},
{
"epoch": 0.81,
"learning_rate": 4.164696786726593e-05,
"loss": 0.5348,
"step": 890
},
{
"epoch": 0.82,
"learning_rate": 4.146914134943843e-05,
"loss": 0.5161,
"step": 900
},
{
"epoch": 0.83,
"learning_rate": 4.1289830353131974e-05,
"loss": 0.5126,
"step": 910
},
{
"epoch": 0.83,
"learning_rate": 4.110905104089621e-05,
"loss": 0.5608,
"step": 920
},
{
"epoch": 0.84,
"learning_rate": 4.0926819707630346e-05,
"loss": 0.5306,
"step": 930
},
{
"epoch": 0.85,
"learning_rate": 4.074315277911438e-05,
"loss": 0.5399,
"step": 940
},
{
"epoch": 0.86,
"learning_rate": 4.055806681052852e-05,
"loss": 0.5085,
"step": 950
},
{
"epoch": 0.87,
"learning_rate": 4.0371578484960906e-05,
"loss": 0.528,
"step": 960
},
{
"epoch": 0.88,
"learning_rate": 4.018370461190392e-05,
"loss": 0.5315,
"step": 970
},
{
"epoch": 0.89,
"learning_rate": 3.9994462125739e-05,
"loss": 0.5357,
"step": 980
},
{
"epoch": 0.9,
"learning_rate": 3.980386808421018e-05,
"loss": 0.5262,
"step": 990
},
{
"epoch": 0.91,
"learning_rate": 3.961193966688664e-05,
"loss": 0.5546,
"step": 1000
},
{
"epoch": 0.92,
"learning_rate": 3.941869417361413e-05,
"loss": 0.5125,
"step": 1010
},
{
"epoch": 0.92,
"learning_rate": 3.922414902295562e-05,
"loss": 0.5429,
"step": 1020
},
{
"epoch": 0.93,
"learning_rate": 3.90283217506213e-05,
"loss": 0.5559,
"step": 1030
},
{
"epoch": 0.94,
"learning_rate": 3.883123000788787e-05,
"loss": 0.5147,
"step": 1040
},
{
"epoch": 0.95,
"learning_rate": 3.86328915600076e-05,
"loss": 0.5537,
"step": 1050
},
{
"epoch": 0.96,
"learning_rate": 3.843332428460696e-05,
"loss": 0.4949,
"step": 1060
},
{
"epoch": 0.97,
"learning_rate": 3.823254617007522e-05,
"loss": 0.514,
"step": 1070
},
{
"epoch": 0.98,
"learning_rate": 3.803057531394302e-05,
"loss": 0.5361,
"step": 1080
},
{
"epoch": 0.99,
"learning_rate": 3.782742992125108e-05,
"loss": 0.5687,
"step": 1090
},
{
"epoch": 1.0,
"learning_rate": 3.762312830290932e-05,
"loss": 0.5287,
"step": 1100
},
{
"epoch": 1.01,
"learning_rate": 3.741768887404634e-05,
"loss": 0.4718,
"step": 1110
},
{
"epoch": 1.02,
"learning_rate": 3.7211130152349496e-05,
"loss": 0.5207,
"step": 1120
},
{
"epoch": 1.02,
"learning_rate": 3.7003470756395845e-05,
"loss": 0.4424,
"step": 1130
},
{
"epoch": 1.03,
"learning_rate": 3.679472940397387e-05,
"loss": 0.5109,
"step": 1140
},
{
"epoch": 1.04,
"learning_rate": 3.658492491039631e-05,
"loss": 0.4586,
"step": 1150
},
{
"epoch": 1.05,
"learning_rate": 3.6374076186804276e-05,
"loss": 0.4981,
"step": 1160
},
{
"epoch": 1.06,
"learning_rate": 3.6162202238462546e-05,
"loss": 0.4991,
"step": 1170
},
{
"epoch": 1.07,
"learning_rate": 3.594932216304658e-05,
"loss": 0.5041,
"step": 1180
},
{
"epoch": 1.08,
"learning_rate": 3.573545514892109e-05,
"loss": 0.4927,
"step": 1190
},
{
"epoch": 1.09,
"learning_rate": 3.552062047341045e-05,
"loss": 0.4954,
"step": 1200
},
{
"epoch": 1.1,
"learning_rate": 3.5304837501061075e-05,
"loss": 0.4924,
"step": 1210
},
{
"epoch": 1.11,
"learning_rate": 3.5088125681896e-05,
"loss": 0.5002,
"step": 1220
},
{
"epoch": 1.12,
"learning_rate": 3.487050454966171e-05,
"loss": 0.5133,
"step": 1230
},
{
"epoch": 1.12,
"learning_rate": 3.465199372006738e-05,
"loss": 0.505,
"step": 1240
},
{
"epoch": 1.13,
"learning_rate": 3.443261288901683e-05,
"loss": 0.5264,
"step": 1250
},
{
"epoch": 1.14,
"learning_rate": 3.4212381830833184e-05,
"loss": 0.5201,
"step": 1260
},
{
"epoch": 1.15,
"learning_rate": 3.399132039647644e-05,
"loss": 0.4936,
"step": 1270
},
{
"epoch": 1.16,
"learning_rate": 3.3769448511754186e-05,
"loss": 0.4969,
"step": 1280
},
{
"epoch": 1.17,
"learning_rate": 3.3546786175525566e-05,
"loss": 0.4781,
"step": 1290
},
{
"epoch": 1.18,
"learning_rate": 3.332335345789863e-05,
"loss": 0.5155,
"step": 1300
},
{
"epoch": 1.19,
"learning_rate": 3.309917049842126e-05,
"loss": 0.5064,
"step": 1310
},
{
"epoch": 1.2,
"learning_rate": 3.287425750426585e-05,
"loss": 0.4699,
"step": 1320
},
{
"epoch": 1.21,
"learning_rate": 3.2648634748407945e-05,
"loss": 0.5182,
"step": 1330
},
{
"epoch": 1.21,
"learning_rate": 3.2422322567798815e-05,
"loss": 0.4881,
"step": 1340
},
{
"epoch": 1.22,
"learning_rate": 3.2195341361532424e-05,
"loss": 0.4931,
"step": 1350
},
{
"epoch": 1.23,
"learning_rate": 3.196771158900666e-05,
"loss": 0.524,
"step": 1360
},
{
"epoch": 1.24,
"learning_rate": 3.173945376807921e-05,
"loss": 0.5238,
"step": 1370
},
{
"epoch": 1.25,
"learning_rate": 3.151058847321815e-05,
"loss": 0.4506,
"step": 1380
},
{
"epoch": 1.26,
"learning_rate": 3.1281136333647416e-05,
"loss": 0.4688,
"step": 1390
},
{
"epoch": 1.27,
"learning_rate": 3.1051118031487326e-05,
"loss": 0.4495,
"step": 1400
},
{
"epoch": 1.28,
"learning_rate": 3.0820554299890384e-05,
"loss": 0.4999,
"step": 1410
},
{
"epoch": 1.29,
"learning_rate": 3.058946592117248e-05,
"loss": 0.513,
"step": 1420
},
{
"epoch": 1.3,
"learning_rate": 3.035787372493956e-05,
"loss": 0.4841,
"step": 1430
},
{
"epoch": 1.31,
"learning_rate": 3.012579858621019e-05,
"loss": 0.4783,
"step": 1440
},
{
"epoch": 1.31,
"learning_rate": 2.9893261423533893e-05,
"loss": 0.4896,
"step": 1450
},
{
"epoch": 1.32,
"learning_rate": 2.9660283197105632e-05,
"loss": 0.4491,
"step": 1460
},
{
"epoch": 1.33,
"learning_rate": 2.9426884906876517e-05,
"loss": 0.5361,
"step": 1470
},
{
"epoch": 1.34,
"learning_rate": 2.9193087590660944e-05,
"loss": 0.4871,
"step": 1480
},
{
"epoch": 1.35,
"learning_rate": 2.8958912322240282e-05,
"loss": 0.4915,
"step": 1490
},
{
"epoch": 1.36,
"learning_rate": 2.87243802094634e-05,
"loss": 0.4772,
"step": 1500
},
{
"epoch": 1.37,
"learning_rate": 2.8489512392343993e-05,
"loss": 0.4973,
"step": 1510
},
{
"epoch": 1.38,
"learning_rate": 2.8254330041155163e-05,
"loss": 0.4987,
"step": 1520
},
{
"epoch": 1.39,
"learning_rate": 2.8018854354521152e-05,
"loss": 0.489,
"step": 1530
},
{
"epoch": 1.4,
"learning_rate": 2.778310655750656e-05,
"loss": 0.4624,
"step": 1540
},
{
"epoch": 1.41,
"learning_rate": 2.7547107899703196e-05,
"loss": 0.5009,
"step": 1550
},
{
"epoch": 1.41,
"learning_rate": 2.7310879653314702e-05,
"loss": 0.5004,
"step": 1560
},
{
"epoch": 1.42,
"learning_rate": 2.7074443111239133e-05,
"loss": 0.4825,
"step": 1570
},
{
"epoch": 1.43,
"learning_rate": 2.6837819585149697e-05,
"loss": 0.4262,
"step": 1580
},
{
"epoch": 1.44,
"learning_rate": 2.660103040357376e-05,
"loss": 0.4871,
"step": 1590
},
{
"epoch": 1.45,
"learning_rate": 2.636409690997039e-05,
"loss": 0.4874,
"step": 1600
},
{
"epoch": 1.46,
"learning_rate": 2.6127040460806495e-05,
"loss": 0.49,
"step": 1610
},
{
"epoch": 1.47,
"learning_rate": 2.5889882423631817e-05,
"loss": 0.5357,
"step": 1620
},
{
"epoch": 1.48,
"learning_rate": 2.5652644175152928e-05,
"loss": 0.499,
"step": 1630
},
{
"epoch": 1.49,
"learning_rate": 2.5415347099306436e-05,
"loss": 0.5093,
"step": 1640
},
{
"epoch": 1.5,
"learning_rate": 2.5178012585331433e-05,
"loss": 0.4934,
"step": 1650
},
{
"epoch": 1.5,
"learning_rate": 2.4940662025841592e-05,
"loss": 0.5063,
"step": 1660
},
{
"epoch": 1.51,
"learning_rate": 2.470331681489686e-05,
"loss": 0.4538,
"step": 1670
},
{
"epoch": 1.52,
"learning_rate": 2.4465998346075128e-05,
"loss": 0.4856,
"step": 1680
},
{
"epoch": 1.53,
"learning_rate": 2.422872801054378e-05,
"loss": 0.4774,
"step": 1690
},
{
"epoch": 1.54,
"learning_rate": 2.3991527195131627e-05,
"loss": 0.537,
"step": 1700
},
{
"epoch": 1.55,
"learning_rate": 2.375441728040117e-05,
"loss": 0.4859,
"step": 1710
},
{
"epoch": 1.56,
"learning_rate": 2.3517419638721357e-05,
"loss": 0.5388,
"step": 1720
},
{
"epoch": 1.57,
"learning_rate": 2.3280555632341238e-05,
"loss": 0.5142,
"step": 1730
},
{
"epoch": 1.58,
"learning_rate": 2.3043846611464316e-05,
"loss": 0.4888,
"step": 1740
},
{
"epoch": 1.59,
"learning_rate": 2.2807313912324242e-05,
"loss": 0.4642,
"step": 1750
},
{
"epoch": 1.6,
"learning_rate": 2.2570978855261507e-05,
"loss": 0.5062,
"step": 1760
},
{
"epoch": 1.6,
"learning_rate": 2.2334862742801756e-05,
"loss": 0.4726,
"step": 1770
},
{
"epoch": 1.61,
"learning_rate": 2.2098986857735644e-05,
"loss": 0.53,
"step": 1780
},
{
"epoch": 1.62,
"learning_rate": 2.186337246120045e-05,
"loss": 0.5003,
"step": 1790
},
{
"epoch": 1.63,
"learning_rate": 2.1628040790763667e-05,
"loss": 0.4687,
"step": 1800
},
{
"epoch": 1.64,
"learning_rate": 2.1393013058508708e-05,
"loss": 0.4609,
"step": 1810
},
{
"epoch": 1.65,
"learning_rate": 2.1158310449122948e-05,
"loss": 0.5066,
"step": 1820
},
{
"epoch": 1.66,
"learning_rate": 2.092395411798813e-05,
"loss": 0.4846,
"step": 1830
},
{
"epoch": 1.67,
"learning_rate": 2.0689965189273572e-05,
"loss": 0.5068,
"step": 1840
},
{
"epoch": 1.68,
"learning_rate": 2.045636475403201e-05,
"loss": 0.4671,
"step": 1850
},
{
"epoch": 1.69,
"learning_rate": 2.022317386829857e-05,
"loss": 0.4876,
"step": 1860
},
{
"epoch": 1.7,
"learning_rate": 1.9990413551192843e-05,
"loss": 0.501,
"step": 1870
},
{
"epoch": 1.7,
"learning_rate": 1.9758104783024237e-05,
"loss": 0.507,
"step": 1880
},
{
"epoch": 1.71,
"learning_rate": 1.9526268503400946e-05,
"loss": 0.4716,
"step": 1890
},
{
"epoch": 1.72,
"learning_rate": 1.9294925609342443e-05,
"loss": 0.4714,
"step": 1900
},
{
"epoch": 1.73,
"learning_rate": 1.9064096953395954e-05,
"loss": 0.5107,
"step": 1910
},
{
"epoch": 1.74,
"learning_rate": 1.8833803341756816e-05,
"loss": 0.4775,
"step": 1920
},
{
"epoch": 1.75,
"learning_rate": 1.8604065532393094e-05,
"loss": 0.4835,
"step": 1930
},
{
"epoch": 1.76,
"learning_rate": 1.8374904233174543e-05,
"loss": 0.5042,
"step": 1940
},
{
"epoch": 1.77,
"learning_rate": 1.8146340100006014e-05,
"loss": 0.452,
"step": 1950
},
{
"epoch": 1.78,
"learning_rate": 1.7918393734965645e-05,
"loss": 0.5025,
"step": 1960
},
{
"epoch": 1.79,
"learning_rate": 1.7691085684447802e-05,
"loss": 0.4923,
"step": 1970
},
{
"epoch": 1.8,
"learning_rate": 1.746443643731115e-05,
"loss": 0.4653,
"step": 1980
},
{
"epoch": 1.8,
"learning_rate": 1.723846642303177e-05,
"loss": 0.4577,
"step": 1990
},
{
"epoch": 1.81,
"learning_rate": 1.701319600986179e-05,
"loss": 0.5044,
"step": 2000
},
{
"epoch": 1.82,
"learning_rate": 1.6788645502993396e-05,
"loss": 0.5008,
"step": 2010
},
{
"epoch": 1.83,
"learning_rate": 1.65648351427286e-05,
"loss": 0.4531,
"step": 2020
},
{
"epoch": 1.84,
"learning_rate": 1.634178510265486e-05,
"loss": 0.4518,
"step": 2030
},
{
"epoch": 1.85,
"learning_rate": 1.611951548782665e-05,
"loss": 0.4898,
"step": 2040
},
{
"epoch": 1.86,
"learning_rate": 1.589804633295331e-05,
"loss": 0.5001,
"step": 2050
},
{
"epoch": 1.87,
"learning_rate": 1.5677397600593108e-05,
"loss": 0.4747,
"step": 2060
},
{
"epoch": 1.88,
"learning_rate": 1.5457589179353964e-05,
"loss": 0.4872,
"step": 2070
},
{
"epoch": 1.89,
"learning_rate": 1.5238640882100658e-05,
"loss": 0.4893,
"step": 2080
},
{
"epoch": 1.89,
"learning_rate": 1.5020572444169007e-05,
"loss": 0.4987,
"step": 2090
},
{
"epoch": 1.9,
"learning_rate": 1.4803403521586995e-05,
"loss": 0.4679,
"step": 2100
},
{
"epoch": 1.91,
"learning_rate": 1.4587153689302993e-05,
"loss": 0.4659,
"step": 2110
},
{
"epoch": 1.92,
"learning_rate": 1.4371842439421396e-05,
"loss": 0.4961,
"step": 2120
},
{
"epoch": 1.93,
"learning_rate": 1.4157489179445583e-05,
"loss": 0.4935,
"step": 2130
},
{
"epoch": 1.94,
"learning_rate": 1.3944113230528661e-05,
"loss": 0.4561,
"step": 2140
},
{
"epoch": 1.95,
"learning_rate": 1.3731733825731857e-05,
"loss": 0.4755,
"step": 2150
},
{
"epoch": 1.96,
"learning_rate": 1.3520370108290953e-05,
"loss": 0.4951,
"step": 2160
},
{
"epoch": 1.97,
"learning_rate": 1.3310041129890737e-05,
"loss": 0.475,
"step": 2170
},
{
"epoch": 1.98,
"learning_rate": 1.3100765848947753e-05,
"loss": 0.5023,
"step": 2180
},
{
"epoch": 1.99,
"learning_rate": 1.2892563128901468e-05,
"loss": 0.4529,
"step": 2190
},
{
"epoch": 1.99,
"learning_rate": 1.2685451736513942e-05,
"loss": 0.4688,
"step": 2200
},
{
"epoch": 2.0,
"learning_rate": 1.2479450340178284e-05,
"loss": 0.4713,
"step": 2210
},
{
"epoch": 2.01,
"learning_rate": 1.2274577508235899e-05,
"loss": 0.4398,
"step": 2220
},
{
"epoch": 2.02,
"learning_rate": 1.2070851707302854e-05,
"loss": 0.4439,
"step": 2230
},
{
"epoch": 2.03,
"learning_rate": 1.1868291300605295e-05,
"loss": 0.469,
"step": 2240
},
{
"epoch": 2.04,
"learning_rate": 1.1666914546324254e-05,
"loss": 0.4142,
"step": 2250
},
{
"epoch": 2.05,
"learning_rate": 1.1466739595949954e-05,
"loss": 0.4307,
"step": 2260
},
{
"epoch": 2.06,
"learning_rate": 1.1267784492645628e-05,
"loss": 0.472,
"step": 2270
},
{
"epoch": 2.07,
"learning_rate": 1.1070067169621215e-05,
"loss": 0.4416,
"step": 2280
},
{
"epoch": 2.08,
"learning_rate": 1.0873605448516864e-05,
"loss": 0.4922,
"step": 2290
},
{
"epoch": 2.09,
"learning_rate": 1.0678417037796607e-05,
"loss": 0.4398,
"step": 2300
},
{
"epoch": 2.09,
"learning_rate": 1.0484519531152099e-05,
"loss": 0.4936,
"step": 2310
},
{
"epoch": 2.1,
"learning_rate": 1.0291930405916853e-05,
"loss": 0.4647,
"step": 2320
},
{
"epoch": 2.11,
"learning_rate": 1.0100667021490823e-05,
"loss": 0.4681,
"step": 2330
},
{
"epoch": 2.12,
"learning_rate": 9.910746617775703e-06,
"loss": 0.4836,
"step": 2340
},
{
"epoch": 2.13,
"learning_rate": 9.722186313621004e-06,
"loss": 0.4718,
"step": 2350
},
{
"epoch": 2.14,
"learning_rate": 9.535003105280955e-06,
"loss": 0.4677,
"step": 2360
},
{
"epoch": 2.15,
"learning_rate": 9.349213864882556e-06,
"loss": 0.4573,
"step": 2370
},
{
"epoch": 2.16,
"learning_rate": 9.164835338904753e-06,
"loss": 0.4483,
"step": 2380
},
{
"epoch": 2.17,
"learning_rate": 8.981884146668995e-06,
"loss": 0.4143,
"step": 2390
},
{
"epoch": 2.18,
"learning_rate": 8.800376778841171e-06,
"loss": 0.4634,
"step": 2400
},
{
"epoch": 2.18,
"learning_rate": 8.62032959594521e-06,
"loss": 0.4403,
"step": 2410
},
{
"epoch": 2.19,
"learning_rate": 8.441758826888421e-06,
"loss": 0.4536,
"step": 2420
},
{
"epoch": 2.2,
"learning_rate": 8.264680567498623e-06,
"loss": 0.4579,
"step": 2430
},
{
"epoch": 2.21,
"learning_rate": 8.089110779073362e-06,
"loss": 0.4465,
"step": 2440
},
{
"epoch": 2.22,
"learning_rate": 7.915065286941148e-06,
"loss": 0.4767,
"step": 2450
},
{
"epoch": 2.23,
"learning_rate": 7.742559779035086e-06,
"loss": 0.4905,
"step": 2460
},
{
"epoch": 2.24,
"learning_rate": 7.5716098044787295e-06,
"loss": 0.4821,
"step": 2470
},
{
"epoch": 2.25,
"learning_rate": 7.402230772184609e-06,
"loss": 0.4678,
"step": 2480
},
{
"epoch": 2.26,
"learning_rate": 7.234437949465256e-06,
"loss": 0.4078,
"step": 2490
},
{
"epoch": 2.27,
"learning_rate": 7.068246460657082e-06,
"loss": 0.4464,
"step": 2500
},
{
"epoch": 2.28,
"learning_rate": 6.903671285757118e-06,
"loss": 0.5072,
"step": 2510
},
{
"epoch": 2.28,
"learning_rate": 6.740727259072768e-06,
"loss": 0.4362,
"step": 2520
},
{
"epoch": 2.29,
"learning_rate": 6.579429067884671e-06,
"loss": 0.4391,
"step": 2530
},
{
"epoch": 2.3,
"learning_rate": 6.419791251122828e-06,
"loss": 0.4321,
"step": 2540
},
{
"epoch": 2.31,
"learning_rate": 6.261828198056157e-06,
"loss": 0.466,
"step": 2550
},
{
"epoch": 2.32,
"learning_rate": 6.105554146995429e-06,
"loss": 0.471,
"step": 2560
},
{
"epoch": 2.33,
"learning_rate": 5.950983184009889e-06,
"loss": 0.449,
"step": 2570
},
{
"epoch": 2.34,
"learning_rate": 5.79812924165761e-06,
"loss": 0.4913,
"step": 2580
},
{
"epoch": 2.35,
"learning_rate": 5.647006097729604e-06,
"loss": 0.4324,
"step": 2590
},
{
"epoch": 2.36,
"learning_rate": 5.49762737400798e-06,
"loss": 0.4746,
"step": 2600
},
{
"epoch": 2.37,
"learning_rate": 5.350006535038079e-06,
"loss": 0.4356,
"step": 2610
},
{
"epoch": 2.38,
"learning_rate": 5.204156886914863e-06,
"loss": 0.4221,
"step": 2620
},
{
"epoch": 2.38,
"learning_rate": 5.060091576083487e-06,
"loss": 0.4374,
"step": 2630
},
{
"epoch": 2.39,
"learning_rate": 4.9178235881543855e-06,
"loss": 0.426,
"step": 2640
},
{
"epoch": 2.4,
"learning_rate": 4.7773657467327346e-06,
"loss": 0.4299,
"step": 2650
},
{
"epoch": 2.41,
"learning_rate": 4.638730712262593e-06,
"loss": 0.4551,
"step": 2660
},
{
"epoch": 2.42,
"learning_rate": 4.50193098088573e-06,
"loss": 0.4554,
"step": 2670
},
{
"epoch": 2.43,
"learning_rate": 4.366978883315248e-06,
"loss": 0.4505,
"step": 2680
},
{
"epoch": 2.44,
"learning_rate": 4.233886583724162e-06,
"loss": 0.4837,
"step": 2690
},
{
"epoch": 2.45,
"learning_rate": 4.1026660786489065e-06,
"loss": 0.4772,
"step": 2700
},
{
"epoch": 2.46,
"learning_rate": 3.97332919590806e-06,
"loss": 0.4122,
"step": 2710
},
{
"epoch": 2.47,
"learning_rate": 3.8458875935361686e-06,
"loss": 0.4439,
"step": 2720
},
{
"epoch": 2.48,
"learning_rate": 3.720352758732956e-06,
"loss": 0.4806,
"step": 2730
},
{
"epoch": 2.48,
"learning_rate": 3.5967360068279164e-06,
"loss": 0.4608,
"step": 2740
},
{
"epoch": 2.49,
"learning_rate": 3.47504848026034e-06,
"loss": 0.4809,
"step": 2750
},
{
"epoch": 2.5,
"learning_rate": 3.355301147575016e-06,
"loss": 0.4484,
"step": 2760
},
{
"epoch": 2.51,
"learning_rate": 3.2375048024335163e-06,
"loss": 0.4982,
"step": 2770
},
{
"epoch": 2.52,
"learning_rate": 3.1216700626413324e-06,
"loss": 0.4707,
"step": 2780
},
{
"epoch": 2.53,
"learning_rate": 3.007807369190768e-06,
"loss": 0.4655,
"step": 2790
},
{
"epoch": 2.54,
"learning_rate": 2.89592698531988e-06,
"loss": 0.4566,
"step": 2800
},
{
"epoch": 2.55,
"learning_rate": 2.7860389955873256e-06,
"loss": 0.4153,
"step": 2810
},
{
"epoch": 2.56,
"learning_rate": 2.6781533049633963e-06,
"loss": 0.4386,
"step": 2820
},
{
"epoch": 2.57,
"learning_rate": 2.5722796379372196e-06,
"loss": 0.4663,
"step": 2830
},
{
"epoch": 2.57,
"learning_rate": 2.4684275376402025e-06,
"loss": 0.4283,
"step": 2840
},
{
"epoch": 2.58,
"learning_rate": 2.3666063649858736e-06,
"loss": 0.4853,
"step": 2850
},
{
"epoch": 2.59,
"learning_rate": 2.2668252978260835e-06,
"loss": 0.4587,
"step": 2860
},
{
"epoch": 2.6,
"learning_rate": 2.169093330123767e-06,
"loss": 0.5012,
"step": 2870
},
{
"epoch": 2.61,
"learning_rate": 2.0734192711422406e-06,
"loss": 0.4482,
"step": 2880
},
{
"epoch": 2.62,
"learning_rate": 1.979811744651161e-06,
"loss": 0.4864,
"step": 2890
},
{
"epoch": 2.63,
"learning_rate": 1.8882791881492278e-06,
"loss": 0.4551,
"step": 2900
},
{
"epoch": 2.64,
"learning_rate": 1.7988298521036172e-06,
"loss": 0.44,
"step": 2910
},
{
"epoch": 2.65,
"learning_rate": 1.7114717992063463e-06,
"loss": 0.4589,
"step": 2920
},
{
"epoch": 2.66,
"learning_rate": 1.6262129036474837e-06,
"loss": 0.4587,
"step": 2930
},
{
"epoch": 2.67,
"learning_rate": 1.543060850405445e-06,
"loss": 0.4385,
"step": 2940
},
{
"epoch": 2.67,
"learning_rate": 1.4620231345542497e-06,
"loss": 0.4963,
"step": 2950
},
{
"epoch": 2.68,
"learning_rate": 1.3831070605879598e-06,
"loss": 0.4926,
"step": 2960
},
{
"epoch": 2.69,
"learning_rate": 1.3063197417622602e-06,
"loss": 0.452,
"step": 2970
},
{
"epoch": 2.7,
"learning_rate": 1.2316680994533174e-06,
"loss": 0.4193,
"step": 2980
},
{
"epoch": 2.71,
"learning_rate": 1.1591588625338767e-06,
"loss": 0.4179,
"step": 2990
},
{
"epoch": 2.72,
"learning_rate": 1.0887985667667522e-06,
"loss": 0.4563,
"step": 3000
}
],
"max_steps": 3309,
"num_train_epochs": 3,
"total_flos": 1.728722097093673e+17,
"trial_name": null,
"trial_params": null
}